Using a natural experimental setting where a solar PV subsidy is assigned randomly to applying households, we estimate the impact of subsidy provision on the adoption of solar PV, installed capacity, timing of the adoption and energy consumption. The results imply that the provision of subsidy leads to an 11.9 percent increase in the probability of adopting solar PV among the households who applied for the subsidy. The findings also indicate that the households who are accepted to the subsidy program invest in an 8.9 percent larger installation as compared to rejected households, and adopt solar panels 55.9 percent faster. Finally, examining the subsequent electricity consumption of the applicants, we report that subsidy provision leads to a 3.74 percent decrease in households' electricity consumption as compared to the rejected applicant group after 1 year, and a 3.93 percent decrease after 5 years.