To the Editor: Dr. Gau and colleagues have raised several issues that we attempted to address in our original paper.1 In response to the initial comment, it is important to state that our study was a randomized trial and that the comparison between the treatment and placebo groups was merely to indicate that balance between groups was achieved by randomization. Having performed more than 20 comparisons between the two groups, it can be expected that the randomized groups would be different in one variable, in this case previous hip fracture, at the P=.05 level of significance. This group's previous falls rates have been well studied and published in a previous paper.2 Because this study was a randomized trial, it is arguable whether adjustment for baseline variables should be performed, particularly for variables for which there was no evidence of differences at baseline, although as stated in our original paper (see page 1885), adjustment for baseline 25-hydroxyvitamin D levels did not substantially alter the effect estimates. There was no follow-up of outside exposure, walking scores, vitamin D status, or parathyroid hormone levels, because these were not outcomes of interest, and resources precluded the gathering of these relatively unimportant data. The study was designed to evaluate the effect of vitamin D supplementation on the clinically important outcomes of falls and fractures, and previous observational studies within the residential care context had already evaluated the relationship between falls, vitamin D levels, parathyroid hormone levels, outdoor exposure, and walking ability.2–4 The trial was not powered to detect differences fracture in rates at individual skeletal sites. As indicated in the paper, there were major concerns regarding the meta-analysis concerning vitamin D supplementation and falls published in JAMA,5 which excluded some well-performed negative studies and was dependent on data of a single study of calcitriol treatment. This meta-analysis is now out of date, because there have been two large negative studies of vitamin D supplementation (unfortunately with poor adherence) performed in community-dwelling older people,6,7 although as we have demonstrated, in relation to falls, and was originally demonstrated8 in relation to fractures, the situation in residential care is different from community-dwelling cohorts. Nearly half of all nursing home residents are severely vitamin D deficient, and nearly all assisted living facility residents and the remainder of the nursing home residents are in the lower half of the reference range and therefore stand to benefit from simple vitamin D supplementation. The screen-and-treat strategy suggested by Gau and colleagues is unlikely to achieve similar benefits in these vulnerable populations. For example, in Australia, only 3% of residents were taking vitamin D supplements.1 This may not be the case elsewhere, but it is probable that, internationally, the majority of aged care residents who would benefit from vitamin D supplements do not receive this treatment in the absence of a policy of universal supplementation. For this reason, we stand by our recommendation that all nursing home and assisted living residents, without specific contraindications, should be supplemented with vitamin D. Financial Disclosure: None of the authors have any perceived financial, or otherwise, conflicts of interest. Author Contributions: All authors contributed to this response letter. Sponsor's Role: None.