Teachers differ greatly in how much they teach their students, but little is known about which teacher attributes account for this. We estimate the causal effect of teacher subject knowledge on student achievement using within-teacher within-student variation, exploiting a unique Peruvian 6th-grade dataset that tested both students and their teachers in two subjects. We circumvent omitted-variable and selection biases using student and teacher fixed effects and observing teachers teaching both subjects in one-classroom-per-grade schools. After measurement-error correction, one standard deviation in subject-specific teacher achievement increases student achievement by about 10 percent of a standard deviation.