Osteogenesis imperfecta (OI) is a heterogeneous disease of connective tissue, the cardinal symptom being fractures and severity ranging from mild to lethal. Dominant mutations in collagen I, encoded by COL1A1 and COL1A2, cause >90% of cases. To delineate genotype-phenotype correlations and pharmaco-genetic response, collagen I was sequenced in 150 unrelated Swedish families and clinical data were collected in Paper I. Mutation type, gene affected, and N- to C-terminal location correlated with phenotype and severity. Bisphosphonate response assessed by calculated yearly change in lumbar spine bone mineral density (BMD) was inversely related to age and BMD at treatment initiation. Mutations associated with a more severe phenotype exhibited an increased response after 2 years; however, all types of OI responded well. To investigate the effect of naturally occurring variations in collagen I, the only common coding single nucleotide polymorphism (rs42524 in COL1A2) was genotyped in 2004 healthy men in Paper II. Heterozygous genotype was associated with decreased BMD and an increased risk of stroke. An adolescent with repeated fractures despite a markedly high BMD harbored a unique C-terminal procollagen cleavage-site mutation in COL1A1, which motivated extensive investigations in concert with a similar COL1A2 case in Paper III. The probands were found to have impaired procollagen processing, incorporation of collagen with retained C-propeptide in matrix and increased mineral to matrix ratio, which demonstrates that C-propeptide cleavage is crucial to normal bone mineralization and structure. Bisphosphonate therapy has insufficient effect in OI, and as classical OI is a dominant disorder severe cases would benefit from silencing of the mutated allele. In Paper IV and V small interfering RNAs (siRNAs) were used to allele-specifically target primary human bone cells heterozygous for I) a coding polymorphism in COL1A2 and II) insertion/deletions in the 3’UTR of COL1A1 and COL1A2. Results were promising with altered allele ratios and decreased mRNA levels in the predicted fashion. To summarize, this thesis found that collagen I is crucial to bone and connective tissue and that collagen I mutations create markedly diverse phenotypes. Age, BMD and pharmaco-genetic effects influence the response to bisphosphonate therapy in individuals with OI; however, novel approaches are needed. Utilizing allele-specific siRNAs may be a way forward in the treatment of severe OI.