This paper presents an approach to making inferences about common activations in groups of subjects using fMRI. It suggests that activations common to all subjects reflect aspects of functional anatomy that may be typical of the population from which the group was sampled. These commonalities can be identified through conjunction analysis, which tests for the joint refutation of multiple null hypotheses (no activation in any subject). Fixed-effect analyses are generally more sensitive than random-effect analyses, but they only apply to the studied subjects. Conjunction analysis using a fixed-effect model allows inferences about both the studied subjects and the population. The second inference depends on a meta-analytic formulation in terms of a confidence region for the proportion of the population showing the effect. This approach retains the sensitivity of fixed-effect analyses when the inference that only a substantial proportion of the population activates is sufficient. The paper discusses the distinction between fixed- and random-effect models in multisubject fMRI studies. Fixed-effect analyses assume each subject contributes the same fixed effect, while random-effect analyses account for subject-to-subject variability. The paper also addresses session-by-contrast interactions, which can arise from physiological differences or instrumentation. Conjunction analyses are introduced as a method to identify jointly significant activations across multiple subjects. The paper describes an empirical example of this approach applied to an fMRI study of visual responses. The results show that conjunction analysis can identify common activations across subjects and provide confidence intervals for the proportion of the population showing the effect. The paper concludes that conjunction analysis is a useful tool for making population inferences about functional anatomy, while acknowledging that random-effect analyses are still necessary in certain cases. The paper also discusses the advantages and limitations of fixed-effect and random-effect analyses in fMRI studies.This paper presents an approach to making inferences about common activations in groups of subjects using fMRI. It suggests that activations common to all subjects reflect aspects of functional anatomy that may be typical of the population from which the group was sampled. These commonalities can be identified through conjunction analysis, which tests for the joint refutation of multiple null hypotheses (no activation in any subject). Fixed-effect analyses are generally more sensitive than random-effect analyses, but they only apply to the studied subjects. Conjunction analysis using a fixed-effect model allows inferences about both the studied subjects and the population. The second inference depends on a meta-analytic formulation in terms of a confidence region for the proportion of the population showing the effect. This approach retains the sensitivity of fixed-effect analyses when the inference that only a substantial proportion of the population activates is sufficient. The paper discusses the distinction between fixed- and random-effect models in multisubject fMRI studies. Fixed-effect analyses assume each subject contributes the same fixed effect, while random-effect analyses account for subject-to-subject variability. The paper also addresses session-by-contrast interactions, which can arise from physiological differences or instrumentation. Conjunction analyses are introduced as a method to identify jointly significant activations across multiple subjects. The paper describes an empirical example of this approach applied to an fMRI study of visual responses. The results show that conjunction analysis can identify common activations across subjects and provide confidence intervals for the proportion of the population showing the effect. The paper concludes that conjunction analysis is a useful tool for making population inferences about functional anatomy, while acknowledging that random-effect analyses are still necessary in certain cases. The paper also discusses the advantages and limitations of fixed-effect and random-effect analyses in fMRI studies.