We present first results from an experimental evaluation of a high dosage math tutoring program implemented in a secondary school for pre-vocational education located in a low-income neighborhood in Amsterdam, the Netherlands. From a pool of 98 students in their first year of secondary school (aged 12-13) 49 were randomly assigned to receive 2-on-1 tutoring. Treated students received one class period of personalized tutoring daily in the first 16 weeks of the school year. Outcome test scores were observed at the end of the first semester and at the end of the school year. We find treatment effects of 0.44 (p<0.01) and 0.72 (p<0.01) control group standard deviations after one semester on a verbal math test (with math word problems) and a nonverbal math test (math without textual context) respectively. The results indicate substantial learning gains. On the nonverbal math test, treated students gained statistically significantly more in one semester than control students gained in the entire school year.