Global In Silico Clinical Trials Market Growth, 2031

In silico clinical trials use computer modelling and simulation to design, optimise and in some cases partly replace elements of studies in people, by representing biology, disease and treatment effects in mathematical and statistical models. The main methods are physiologically based pharmacokinetic modelling, which predicts drug exposure in different populations; population pharmacokinetic and pharmacodynamic models; quantitative systems pharmacology, which simulates disease pathways and drug mechanisms; virtual patient populations and digital twins, which represent individuals or cohorts and can be used to test dosing or device designs; clinical trial simulation to explore design choices, enrolment and endpoints; and synthetic or external control arms built from historical trial and real-world data. In medical devices, finite element and fluid dynamics models of the heart, bones and vessels are used to test designs on virtual patients. Software and services come from Certara with its Simcyp and Phoenix products, Simulations Plus with GastroPlus and DILIsym, Dassault Systemes with its Living Heart project and Medidata, Synopsys following its 2025 completion of the Ansys acquisition, Unlearn, Novadiscovery, InSilicoTrials, Metrum Research Group, Pumas-AI and a number of consultancies and start-ups. Users include pharmaceutical, biotechnology and device companies, contract research organisations, academic groups and regulators. The approach is supported by model-informed drug development programmes at the Food and Drug Administration and by credibility frameworks such as the ASME V&V 40 standard. Models are applied to select doses, predict drug interactions, support paediatric extrapolation, size trials, identify patients likely to respond and reduce the number of participants who receive placebo.

The market now sits at an inflection point shaped by regulatory openness, advances in artificial intelligence and pressure to improve research and development productivity. Developing a new medicine typically takes more than a decade and costs very large sums, with most candidates failing in clinical trials, so even modest gains from better dose selection or trial design have high value. Regulators have issued guidance on model credibility for devices, on externally controlled trials and on artificial intelligence in regulatory decision-making, and the European Medicines Agency qualified a digital-twin-based method for reducing sample size in 2022. Policy attention to reducing animal testing, including the FDA's 2025 roadmap, encourages computational alternatives, while the growth of real-world data and generative models makes virtual cohorts more realistic. Adoption remains uneven because model validation, data quality and regulatory acceptance vary by use case. For researchers, the key variables through 2031 are the breadth of regulatory acceptance, validation of digital twins against real outcomes, access to high-quality patient-level data, integration of machine learning with mechanistic models, cost and skills, and use in rare disease, paediatrics and medical devices where recruitment is hard.

Drivers •Pressure to Improve R&D Productivity and Speed: Late-stage trial failure and rising trial costs push sponsors to find better ways to select doses, endpoints and patient populations before committing to large studies. Simulation can test many design options quickly, predict outcomes under different assumptions and reduce the chance of costly amendments or failures. Smaller trials, shorter timelines and better probability of success translate directly into savings, so pharmaceutical firms invest in pharmacometrics and systems modelling groups and outsource specialist work. These economic incentives support steady growth through 2031.

•Regulatory Encouragement and Ethical Need in Rare Disease, Paediatrics and Devices: Regulators encourage the use of modelling, including the FDA's model-informed drug development paired meeting programme and its guidance on credibility of computational modelling for medical devices finalised in November 2023, and international work on a general guideline for model-informed development. In rare diseases and children, recruiting enough participants is difficult and sometimes unethical for placebo groups, so synthetic controls and extrapolation based on models are attractive. Device makers use virtual patients to test designs under varied anatomies. Such support increases confidence in using simulation evidence in submissions.

Challenges •Model Validation, Data Quality and Regulatory Acceptance: A model is only as good as the biology and data behind it, and demonstrating that a virtual population represents real patients, including rare subgroups, is difficult. Training data may be biased or incomplete, and mechanistic models contain uncertain parameters that are hard to measure. Regulators accept simulation for some questions, such as drug interactions and paediatric dose selection, but remain cautious about replacing human evidence for efficacy. Lack of clear acceptance criteria for new approaches such as digital twins can lead sponsors to limit use to supportive analyses.



•Skills, Cost and Organisational Adoption: Building and qualifying models requires expertise across pharmacology, statistics, biology and software, and such specialists are in short supply. Large simulations need computing resources and careful documentation to meet regulatory standards, and models developed for one drug may not transfer easily to another. Smaller companies may lack budgets, while large companies face internal barriers between modelling and clinical teams. Data access constraints, privacy rules and intellectual property concerns on shared datasets also slow progress.

Trends •Digital Twins and Generative AI for Virtual Patients: Companies are using machine learning to create digital twins that forecast a patient's outcome under standard care, and use this as a covariate to reduce variance and the number of participants needed, as with Unlearn's PROCOVA method, which received a qualification opinion from the European Medicines Agency in 2022. Generative models can create synthetic patient records that preserve statistical properties without exposing individuals, and large language models are used to extract data from literature and records. Mechanistic and AI approaches are being combined so that models are both interpretable and data-driven. Validation and transparency are key to acceptance.

•Cloud Platforms, Regulatory-Grade Workflows and Device Applications: Providers are moving simulation software to cloud platforms that provide scalable compute, collaboration and audit trails, so that sponsors and regulators can review and reproduce analyses. Libraries of validated models, such as population models for physiology and disease, reduce build time. In medical devices, companies including Dassault and Synopsys through Ansys offer simulation of cardiovascular and orthopaedic devices, and virtual cohorts are used to test performance across patient anatomies. Integration with real-world data and electronic health records allows models to be calibrated and updated, and some regulators are exploring shared evaluation frameworks.

Physiologically based pharmacokinetic and pharmacometric modelling leads by established use because it is widely accepted by regulators for dose and drug interaction questions, while virtual patient and digital twin approaches are the fastest-growing as AI improves and trial sponsors seek efficiency. Physiologically based pharmacokinetic modelling predicts drug concentrations in organs and in special populations such as children, older adults and patients with organ impairment, and is used in submissions to inform labels, with Certara's Simcyp and Simulations Plus's GastroPlus as major tools. Population pharmacokinetic and pharmacodynamic models use trial data to describe variability and support dose selection and exposure-response analysis. Quantitative systems pharmacology models simulate disease networks and drug effects for complex conditions such as immunology, oncology and metabolic disease. Virtual patient and digital twin methods simulate individual or cohort responses, and trial simulation tests design and operating scenarios. Synthetic and external control arms use historical trial data or real-world data, as supplied by Medidata and others, to supplement or replace control groups. Through 2031, hybrid methods that combine mechanistic and machine learning elements are expected to expand.


Phase II and Phase III trials lead as a use because design and population choices have the largest cost and risk at these stages, while early clinical phases and preclinical work are heavy users of pharmacokinetic modelling and post-market applications are growing. In early phases, modelling guides first-in-human dose selection, predicts drug interactions and food effects, and supports decisions to proceed, using preclinical and early clinical data. In Phase II, simulation helps choose doses, endpoints and enrichment strategies, and tests adaptive designs. In Phase III, trial simulation supports sample size, interim analysis plans and patient selection, and synthetic controls may supplement concurrent controls in rare diseases. Post-approval, models support labelling updates, paediatric extension, special populations and real-world effectiveness comparisons. Device studies use virtual cohorts at design and verification stages. Through 2031, growth is expected in Phase II and III uses, in line with expanded regulatory comfort and economic return.


Oncology leads by therapeutic area because it has the largest and most complex pipelines and heavy use of dose optimisation, while neurology, cardiovascular and metabolic disease, infectious disease and immunology, and rare diseases are important and growing areas. Oncology uses modelling for dose and schedule selection of targeted drugs, immunotherapies and combinations, tumour growth models and exposure-response analysis, and the FDA's Project Optimus encourages dose optimisation that models can support. Neurology and central nervous system uses include models of disease progression for Alzheimer's and Parkinson's disease and receptor occupancy simulation. Cardiovascular and metabolic applications include heart simulation for device design and models for diabetes and obesity drugs. Infectious disease and vaccines use transmission and immunity models, and immunology uses systems models of cytokine networks. Rare disease and paediatric programmes use extrapolation and synthetic controls to overcome small populations. Through 2031, rare disease and paediatrics are expected to expand quickly because of the practical need for alternatives to large trials.


Pharmaceutical and biotechnology companies lead as end users because they fund most development programmes and employ modelling teams, while contract research organisations and consultancies deliver outsourced modelling, medical device companies are an expanding group, and regulators and academic institutions are smaller users with influence on standards. Large pharmaceutical firms have in-house pharmacometrics and systems pharmacology groups and license commercial software, using external partners for peaks and specialised models. Biotechnology companies lack such teams and rely on consultancies and contract research organisations, which offer modelling alongside clinical trial services. Medical device and diagnostics companies use simulation to support design verification and regulatory submissions, particularly in cardiovascular, orthopaedic and neuromodulation devices. Regulatory agencies run internal modelling groups that review submissions and develop methods. Academic and non-profit groups, including consortia such as the Simcyp Consortium and virtual human twin initiatives, develop open models and training. Through 2031, outsourcing and software subscriptions are expected to broaden access beyond large companies.


The United States is the largest market, with companies such as Certara, Simulations Plus, Dassault Systemes through Medidata, Synopsys and a range of start-ups, and with major pharmaceutical and device companies adopting modelling. The Food and Drug Administration has a long history of accepting simulation, including the 2008 acceptance of a type 1 diabetes simulator in place of some animal studies, a model-informed drug development paired meeting programme, final guidance on credibility of computational modelling for medical devices in November 2023, draft guidance on externally controlled trials in February 2023 and draft guidance on the use of artificial intelligence in regulatory decision-making in January 2025. The agency's April 2025 plan to reduce animal testing added momentum. Canada has research groups and consultancies in modelling, and Mexico has limited use mostly through multinational sponsors and contract research organisations.

Europe has a strong academic tradition in pharmacometrics and systems medicine, and companies such as Novadiscovery in France, InSilicoTrials in Italy and many consultancies in the United Kingdom, Netherlands, Sweden and Germany. The European Medicines Agency offers scientific advice and qualification opinions for novel methodologies, including its 2022 opinion on a digital twin-based covariate adjustment method, and published a reflection paper on artificial intelligence in the medicine lifecycle in September 2024. The European Commission supports virtual human twin work through initiatives such as EDITH, and industry groups such as the Avicenna Alliance promote regulatory acceptance. The Medical Device Regulation encourages clinical evidence but also allows model-based evidence in some contexts. Data protection rules affect the use of patient-level data, but a collaborative research environment supports innovation.

Asia-Pacific is growing as regulators and sponsors adopt model-informed approaches. Japan's Pharmaceuticals and Medical Devices Agency has pioneered the use of modelling in review and promotes model-informed development, and Japanese companies use pharmacometrics widely. China's regulator has issued guidance on population pharmacokinetics and has increased scrutiny and acceptance of modelling as domestic innovation grows, with local software and consultancies emerging. India's contract research sector provides modelling services to global sponsors, and South Korea has growing pharmacometrics groups. Australia has leading academic groups in pharmacology and a role in early-phase trials, and Singapore supports translational research. Large patient populations, increasing domestic drug development and regulatory alignment with international standards support growth.

Latin America has limited but emerging use of in silico trial methods. Brazil and Mexico have universities with pharmacology and computational biology programmes and multinational sponsors that use modelling for global trials, while local regulators such as ANVISA are aligning guidance with international approaches. Contract research organisations provide some pharmacokinetic analysis for bioequivalence studies. Constraints include limited funding, scarce specialists and less access to large datasets, though growing digital health data and academic collaboration may enable local development. Adoption is likely to be led by multinational sponsors and by use in generic and biosimilar development, with gradual expansion through 2031.

The Middle East and Africa are at an early stage, with interest from governments building clinical research and digital health capacity. The Gulf states, including Saudi Arabia, the United Arab Emirates and Qatar, invest in genomics, artificial intelligence and clinical trial infrastructure as part of diversification programmes, and some hospitals and research centres partner with global modelling firms. Israel has a strong ecosystem of digital health and clinical trial technology companies, including those focused on trial simulation. In Africa, modelling is mainly used in academic research on infectious disease, such as malaria and tuberculosis, and by international collaborations. Data availability, skills and funding limit adoption, but training programmes and partnerships are expected to build capacity through 2031.

July 2025: Synopsys completes acquisition of Ansys: The combination of electronic design and engineering simulation companies strengthened simulation capabilities that are also used for medical device and physiological modelling.

April 2025: FDA announces roadmap to reduce animal testing: The plan encouraged the use of new approach methodologies, including computational models and in vitro systems, in preclinical safety assessment, supporting demand for in silico tools.

January 2025: FDA issues draft guidance on AI in regulatory decision-making: The draft set out a risk-based credibility framework for AI models used to support decisions on drug safety, effectiveness and quality, relevant to digital twins and virtual cohorts.

September 2024: EMA publishes reflection paper on AI in the medicine lifecycle: The paper described expectations for the use of artificial intelligence and models across development, manufacturing and monitoring, offering a framework for sponsors.

November 2023: FDA finalises guidance on credibility of computational modelling for medical devices: The guidance described how to assess and document the credibility of simulations used in device submissions, providing a clear path for virtual testing.


Considered in this report
• Historic Year: 2020
• Base Year: 2025
• Estimated Year: 2026
• Forecast Year: 2031

Aspects covered in this report
• Global In Silico Clinical Trials Market with its segments and regional outlook
• Various drivers and challenges
• Ongoing trends and developments
• Key developments and regulatory milestones

By Modelling Approach
• Physiologically Based Pharmacokinetic and Pharmacometric Modelling
• Quantitative Systems Pharmacology
• Virtual Patients and Digital Twins
• Clinical Trial Simulation
• Synthetic and External Control Arms

By Trial Phase
• Preclinical and Phase I
• Phase II
• Phase III
• Phase IV and Post-Market

By Therapeutic Area
• Oncology
• Neurology
• Cardiovascular and Metabolic Disease
• Infectious Disease and Immunology
• are Disease and Others

By End User
• Pharmaceutical and Biotechnology Companies
• Contract Research Organisations and Consultancies
• Medical Device Companies
• Regulators and Academic Institutions