Indoor residual spraying (IRS) and insecticide-treated bed nets (ITNs) are cornerstone malaria prevention and control methods in Madagascar. From 2016 to 2020, non-pyrethroid IRS was deployed to complement standard pyrethroid ITNs in 14 districts with high malaria burden, targeting 5 to 9 districts each year. Districts received IRS for 1 to 3 consecutive years during the study period. This retrospective observational study uses routine data to evaluate the impacts of IRS overall, sustained IRS over multiple years, and achieving high (≥85%) IRS coverage (structures sprayed/found). We fit a multilevel mixed effects model to data from all 114 districts of Madagascar from July 2016 to June 2021. We estimated the effect of IRS exposure, consecutive years of IRS, and high IRS coverage on monthly population-adjusted RDT-confirmed malaria cases at health facility level. Facilities missing data, and communes missing geolocations were excluded, leaving 84% of records included. The model controlled for ITN survivorship, mass drug administration (MDA), precipitation, enhanced vegetation index (EVI), month, year, and district. Using the fitted model we predicted malaria cases under observed and no IRS scenarios and estimated the number of cases averted by IRS. IRS was associated with reduced case incidence and an estimated 196,075 (79,879-316,809) cases were averted in targeted districts (~15% of the 1.3m reported cases). The effect varied by district and was associated with ITN survivorship, MDA, precipitation, EVI, month and year. One year of IRS was associated with higher incidence versus two (IRR = 1.15, 95%CI = 1.03-1.29) or three (IRR = 1.16, 95%CI = 1.01-1.33). High coverage (achieved in 94% of IRS areas) was associated with a 12% lower incidence rate (IRR=0.88, CI=0.82-0.95) compared to areas with lower coverage. This study suggests that IRS together with ITNs may substantially reduce malaria incidence over ITNs alone, and high spray coverage and >1 year of IRS may confer additional benefits. This work highlights the value of routine data to evaluate the impact of intervention combinations and to inform future targeting decisions in Madagascar. (French)