The beta doses produced by 90Sr-Y and 204 Tl beta sources were determined using three methods: Monte-Carlo calculations, measurements with TLDs, and measurements with an extrapolation chamber. Excellent agreement was obtained by all three methods, except a TLD nonlinear response to beta s was observed, which gives doses approximately 20% high for the 90Sr-Y source and 5% low for the 204Tl source. Also, analyses performed with low-energy beta s using these methods can determine errors in shield thickness covering TLD elements. Direct measurement of skin dose is not possible by the TLDs because the minimum shield thickness for the elements is 13 mg cm-2. A thinner shield for the elements must be used or the data must be extrapolated. Presently, thinner shields for TLD elements are not available, and the thick shields can lead to significant errors in skin dose when exposed to low-energy beta s.