We first examined the effects of a year-long professional development (PD) programme for elementary science teachers on fifth grade student performance on state-mandated science achievement tests of students from a treatment and a comparison group of teachers in the 2009-2010 academic year. Then, we investigated the longer-term impacts by comparing the 2010-2011 student test results of the teachers one year after receiving treatment in 2009-2010 with the students of teachers who received treatment during 2010-2011. Test scores were analysed using a propensity score matching method to examine the relationship between the PD and student achievement. Results showed that even though the treatment teachers were out of the classroom 20% of the school year to attend the PD, there was no difference between their students' science achievement scores and those of the comparison teachers who were in the classroom every day. This is an important finding because many principals and parents are reluctant to provide teachers with release time for PD. We also determined that students of teachers one year after participating in the PD significantly (p < 0.001) with a medium effect size (?[superscript 2] = 0.088) outperformed students of teachers who had just completed the programme. This suggests that it takes time for teachers to implement new teaching strategies and that to observe the impact of an intervention programme, it may be important to expand the timeframe of the programme evaluation.