Can psychoanalytic theory be scientific? Psychoanalysis has a distinct method that has generated a distinct body of evidence, the clinical data. To qualify as a scientific psychology, these data must be able to evidentially support theoretical claims. I discuss three key objections to the methods of psychoanalysis, 'suggestion' (which potentially biases the data), whether the data can be used to provide genuine tests of theoretical hypotheses, and whether we can reliably infer motives from the behavior noted in the clinical data. I conclude that parts of psychoanalytic theory, namely, the psychodynamic model of the mind, can be defended on the basis of the clinical data, but other parts cannot.