Olefin metathesis is an exceptionally versatile and general methodology for the catalytic assembly of carbon-carbon bonds. Ruthenium metathesis catalysts have been widely embraced in academia, and are starting to see industrial uptake. However, the challenges of reliability, catalyst productivity, and catalyst cost have limited implementation even in value-added technology areas such as pharmaceutical manufacturing. Key to the broader adoption of metathesis methodologies is improved understanding of catalyst decomposition. Many studies have focused on phenomenological relationships that relate catalyst activity to substrate structure, and on the synthesis of new catalysts that offer improved activity. Until recently, however, relatively little attention was paid to catalyst decomposition. The first part of this thesis explores a largely overlooked decomposition pathway for “second-generation” olefin metathesis catalysts bearing an N-heterocyclic carbenes (NHC) ligand, with a particular focus on identifying the Ru decomposition products. Efforts directed at the deliberate synthesis of these products led to the discovery of a succinct, high-yielding route to the second-generation catalysts. Multiple reports, including a series of detailed mechanistic studies from our group, have documented the negative impact of phosphine ligands in Ru-catalyzed olefin metathesis. Phosphine-free derivatives are now becoming widely adopted, particularly in pharma, as recognition of these limitations has grown. Decomposition of the phosphine-free catalysts, however, was little explored at the outset of this work. The only documented pathway for intrinsic decomposition (i.e. in the absence of an external agent) was -hydride elimination of the metallacyclobutane (MCB) ring as propene. An alternative mechanism, well established for group 3-7 and first-generation ruthenium metathesis catalysts, is bimolecular coupling (BMC) of the four-coordinate methylidene intermediate. However, this pathway was widely viewed as irrelevant to decomposition of second-generation Ru catalysts. This thesis work complements parallel studies from the Fogg group, which set out to examine the relevance and extent of BMC for this important class of catalysts. First, -hydride elimination was quantified, to assess the importance of the accepted pathway. Even at low catalyst concentrations (2 mM Ru), less than 50% decomposition was shown to arise from -hydride elimination. Parallel studies by Gwen Bailey demonstrated ca. 80% BMC for the fast-initiating catalyst RuCl2H2IMes(=CHPh)(py)2 GIII. Second, the ruthenium products of decomposition were isolated and characterized. Importantly, and in contrast to inferences drawn from the serendipitous isolation of crystalline byproducts (which commonly show a cyclometallated NHC ligand), these complexes show an intact H2IMes group. This rules out NHC activation as central to catalyst decomposition, suggesting that catalyst redesign should not focus on NHC cyclometallation as a core problem. Building on historical observations, precautions against bimolecular coupling are proposed to guide catalyst choice, redesign, and experimental setup. The second part of this thesis work focused on the need for more efficient routes to second-generation Ru metathesis catalysts, and indeed a general lack of convenient, well-behaved precursors to RuCl2(H2IMes). This challenge was met by building on early studies in which metathesis catalysts were generated in situ by thermal or photochemical activation of RuCl2(p-cymene)(PCy3) in the presence of diazoesters. Such piano-stool complexes (including the IMes analogue) have also been applied more broadly as catalysts, inorganic drugs, sensors, and supramolecular building blocks. However, RuCl2(p-cymene)(H2IMes), which should in principle offer access to the RuCl2(H2IMes) building block, has been described as too unstable for practical use. The basis of the instability of RuCl2(p-cymene)(H2IMes) toward loss of the p-cymene ring was examined. Key factors included control over reaction stoichiometry (i.e. limiting the proportion of the free NHC), limiting exposure to light, and maintaining low concentrations to inhibit bimolecular displacement of the p-cymene ring. A near-quantitative route to RuCl2(p-cymene)(H2IMes) was achieved using appropriate dilutions and rates of reagent addition, and taking precautions against photodecomposition. This approach was used to develop atom-economical syntheses of the Hoveyda catalyst, RuCl2(H2IMes)(=CHAr) (Ar = 2-isopropoxybenzylidene) and RuCl2(H2IMes)(PPh3)(=CHPh), a fast-initiating analogue of GII. Related p-cymene complexes bearing bulky, inflexible imidazolidene or other donors may likewise be accessible.