The hazard of radon-induced lung cancer is not attributed to the radon gas itself but to the short-lived radon progeny. Nevertheless, up to now, there are no widely accepted methods for long-term monitoring of radon progeny with acceptable uncertainty. In the past two decades, methods based on measurements of activities of 210Po implanted in glass objects (or other suitable objects) for retrospective estimation of integrative exposure to radon gas concentrations has been developed. However, estimation of exposure to radon progeny concentrations is more important for the determination of the radon-induced lung cancer risk. In fact, recent research has shown that the 210Po activities deposited in the surfaces of those objects depend on the integrative exposure to short-lived radon progeny concentrations. The present work aims to study in more details the methodology and practice in assessing the retrospective exposure to radon progeny (but not to radon gas) through measuring the activities of implanted 210Po in surfaces of glass objects, which we refer to as retrospective radon progeny dosimetry. Chapter 1 will give an introduction to the background information of the research, which is followed by Chapter 2 which gives a review of the main physics and techniques behind the studies. The (CR–LR) difference technique, based on two different types of solidstate nuclear track detectors, i.e., CR-39 and LR 115, for the determination of implanted 210Po in glass after deposition of short-lived radon progeny, was analyzed in details in Chapter 3. The sensitivities of both detectors were calculated using the Monte Carlo method with V functions particularly derived in previous works of our group for the detectors used in the present experiments. The dependency of the sensitivity ratio on the removed layer of both detectors was determined and verified experimentally. The simulated sensitivity ratios correlated well with the experimental ones. A major finding of the present work was that the sensitivity ratio between the CR-39 and LR 115 detectors depended only weakly on the ratio between the 238U and 232Th concentrations in the glass samples. This is crucial for the application of the (CR–LR) difference technique for retrospective radon exposure assessments, since measurements of the 238U and 232Th concentrations in the relatively small real-life glass samples will make the retrospective radon exposure assessments impractical. The (CR–LR) difference technique for retrospective radon progeny measurements based on implanted 210Po activities in glass objects was then applied in a survey which was described in Chapter 4. In this survey, a total of 48 glass objects in 17 dwellings were examined, but only 19 objects gave results which were sufficiently reliable due to the sensitivity of the method. From these 19 data, an increase in the surface 210Po activities in the glass objects with the age of the glass objects was noticeable as expected. The surface activities of 210Po in the glass objects were then converted to the potential alpha energy concentration (PAEC) through a calibration curve. It was found that the PAEC for dwelling sites did not change significantly with the building age. Chapter 5 will introduce a new method for retrospective radon progeny measurements by applying the so-called “stacked LR 115 detector”. Although the (CR–LR) difference technique has become a default method for retrospective radon progeny measurements, some simplifications or improvements of the methodology are still possible. The proposed stacked LR 115 detector consists of two active layers. The sensitivities of both active layers were calculated using the Monte Carlo method with V functions particularly derived for the LR 115 detectors used in the present experiments. Two glass samples were exposed in an exposure chamber to determine the experimental calibration factors CFg(exp) and CFp(exp) (with the subscripts g and p referring to exposures to radon gas and progeny, respectively), which were then compared with the theoretical calibration factors CFg(sim) and CFp(sim) from simulations through the Jacobi room model. It was encouraging to obtain CFp(exp) ≈ CFp(sim). On the other hand, CFg(exp) >> CFg(sim), which illustrates the superiority of using CFp over CFg. The result that CFp(exp) ≈ CFp(sim) is particularly encouraging since this shows very good agreement between the calibration factor determined experimentally through measurements of 210Po activities on glass objects using the newly proposed stacked LR 115 detectors and that determined theoretically through computer simulations.Based on the concept and the proposed stacked LR 115 detector described in Chapter 5, a mini-survey of contemporary and retrospective radon progeny concentrations at 10 different residential sites was carried out, which is described in Chapter 6. Here, the retrospective values were determined using the stacked LR 115 detectors and the contemporary values were determined using the “proxy-equilibrium factor” method. All measurements using stacked LR 115 detectors on glass objects gave positive values on activities of 210Po implanted in the surface of glass objects. Interestingly, a relationship between contemporary and retrospective radon progeny concentrations was not readily observable. This shows possible changes in the radon exposures over time and highlights the importance of long-term retrospective radon exposure assessments in casecontrol studies to estimate the risk of lung cancer. Chapter 7 will summarize the achievements in the present thesis. All in all, the study has contributed to the field of retrospective radon progeny measurements using solid-state nuclear track detectors.