Neutron attenuation measurements have been used as an instrumental method of analyzing (normally flat) test coupons for the concentration of the 10B nuclide. Calibrated standards of well-characterized 10B content are used to interpret the observed neutron counting rates into the 10B areal density. Recently, there have been challenges to the validity of neutron attenuation measurements and their relationship to criticality safety analyses. For the most part, these challenges have been verbal without any supporting data. The present study was undertaken to provide experimental and analytical investigations of these challenges. The challenges are as follows: 1. It has been claimed that neutrons of any energy (including epithermal and fast neutrons) can be used for attenuation measurements. Spectral and reaction rate calculations are presented to demonstrate that only thermal neutrons have sufficient sensitivity to yield reliable neutron attenuation measurements because of the fundamental 1/v absorption cross section of 10B. 2. It has been alleged that only small-diameter [0.953 cm (3/8 in.)] neutron beams are acceptable for neutron attenuation measurements and that larger-diameter [2.54 cm (1 in.)] beams would "mask" any defects or significant nonuniformities. Both experimental and analytical data are presented to show that the measurements are independent of beam size and that adequate sensitivity to detect any defects or nonuniformities is provided. Criticality calculations are also presented to illustrate that small defects (holes or cracks) have very small effects on results of criticality analyses. 3. It has been postulated that in absorbers using particles of boron carbide, neutrons could stream past discreet particles, reducing the effectiveness of the absorber. While this may be true in attenuation measurements, there is no evidence that neutron streaming has any significant effect in criticality safety analyses. Calculations and an explanation are presented. Neutron attenuation and criticality analyses refer to physically different phenomena with appreciably different path lengths rendering criticality analyses insensitive to streaming.