Sampling is the process of selecting units (e.g., people, organizations, plants) from a population of interest so that by studying the sample, we may fairly generalize our results to the population from which they were chosen. Samples are preferred over a census in scientific research because of the high costs and logistical issues associated with a census. A cornerstone of sampling is a representative sample, which mirrors the population’s characteristics. This chapter explains the various types of populations: target population (could be unknown) and accessible population (sampling unit or the study population). We get our sample from the accessible population, which gives the sampling frame. We have two types of sampling methods: probability (mathematical and random) and non-probability (non-mathematical and non-random). Within the probability sampling: simple random sampling, systematic sampling, stratified sampling, cluster sampling, and multiphase sampling. On the other hand, we have purposive, convenience, theoretical, snowballing, and quota sampling in non-probability sampling. Each of these has benefits and limitations. Purposive sampling has variants such as maximal, critical, typical sampling, and others. Sample size is a critical factor in both quantitative and qualitative research approaches. In quantitative studies, the easiest way to determine sample size is to use an online calculator or formulas and standard tables found in statistics and research method books. As you use these methods, it is recommended that you always select the sample to fit the purpose. Other factors, such as heterogeneity of the population, confidence level and number of aggregations, should be considered. In qualitative studies, the most important is “fit-for-purpose,” which gives rich quality information to fulfil your objectives and answer your research problem. Seven guidelines are provided for selecting a qualitative research sample. A brief final section highlights that the probability and non-probability methods discussed apply to agronomic and livestock studies, specifically biophysical measurements. Sampling in mixed methods research uses probability and non-probability sampling methods.