BACKGROUND: Previous studies demonstrated that fewer mosquitoes enter houses which are screened or have closed eaves. There is little evidence about the effect on malaria infection in humans that changes in house construction may have. This study examines the impact of protective housing improvements on malaria infection on Bioko Island. METHODOLOGY/PRINCIPAL FINDINGS: Data from the annual malaria indicator surveys between 2009 and 2012 were used to assess trends in housing characteristics and their effect on RDT confirmed malaria infection in household members. Odds ratios were adjusted for socio-economic status of the household.22726 children between the ages of 2 and 14 years were tested for P. falciparum. Prevalence of infection in those living in houses with open eaves was 23.0% compared to 18.8% for those living in houses with closed eaves (OR = 0.81, 95% CI 0.67 - 0.98). The prevalence of infection for children in screened houses was 9.1% versus 20.1% for those living in unscreened houses (OR = 0.44, 95% CI 0.27 - 0.71). The proportion of houses with closed eaves increased from 66.0% in 2009 to 74.3% in 2012 (test for trend p = 0.01). The proportion of screened houses remained unchanged over time at 1.3%. CONCLUSION/SIGNIFICANCE: As a malaria control intervention, house modification has the advantages that it is not affected by the growing threat of insecticide resistance; it protects all household members equally and at all times while indoors; and it offers protection against a number of vector borne diseases. The study provides evidence in support of efforts to regulate or encourage housing improvements which impede vector access into residences as part of an integrated vector control approach to complement existing measures which have been only partially successful in reducing malaria transmission in some parts of Bioko.